BibAgent: An Agentic Framework for Traceable Miscitation Detection in Scientific Literature

TL;DR

BibAgent is a scalable, automated framework that verifies citations in scientific literature by integrating retrieval, reasoning, and evidence aggregation, effectively detecting miscitations across disciplines.

Contribution

It introduces BibAgent, a novel end-to-end agentic system with a new Evidence Committee mechanism and a comprehensive MisciteBench dataset for systematic citation verification.

Findings

Outperforms state-of-the-art LLM baselines in accuracy

Effective in both accessible and paywalled sources

Provides transparent detection of citation misalignments

Abstract

Citations are the bedrock of scientific authority, yet their integrity is compromised by widespread miscitations: ranging from nuanced distortions to fabricated references. Systematic citation verification is currently unfeasible; manual review cannot scale to modern publishing volumes, while existing automated tools are restricted by abstract-only analysis or small-scale, domain-specific datasets in part due to the "paywall barrier" of full-text access. We introduce BibAgent, a scalable, end-to-end agentic framework for automated citation verification. BibAgent integrates retrieval, reasoning, and adaptive evidence aggregation, applying distinct strategies for accessible and paywalled sources. For paywalled references, it leverages a novel Evidence Committee mechanism that infers citation validity via downstream citation consensus. To support systematic evaluation, we contribute a 5-category Miscitation Taxonomy and MisciteBench, a massive cross-disciplinary benchmark comprising 6,350 miscitation samples spanning 254 fields. Our results demonstrate that BibAgent outperforms state-of-the-art Large Language Model (LLM) baselines in citation verification accuracy and interpretability, providing scalable, transparent detection of citation misalignments across the scientific literature.